Comprehensive Chronology Of British Turbojet Developments

by John Foster, Jr,
Managing Editor, Aviation

Just-released information an late English turbine-engine types permits presentation of revealing design details, together with genealogy clarifying lines of growth in both centrifugal and axial flow units.

While the great majority of British turbojet engines have been centrifugal-compressor types, stemming from Air Commodore Frank Whittle's original research, the first English studies were made on axial flow compressors.

Current declassification of turbojet information abroad only now reveals this and other facts, making it possible to clearly trace developments and present design and structural details of Britain's more modern types.

Just 20 years ago, Dr A A Griffiths, then head of the Royal Aeronautical Establishment's engine department, presented studies of an aerodynamic theory for a turbine designed for a flow past airfoils rather than through passages. Experimental work along axial-flow lines was continued by the RAE along with British manufacturers, including Metropolitan-Vickers, Armstrong Siddeley, and C A Parsons' organization, up to and through the war.

However, success of the Whittle unit brought concentration on the centrifugal type, as is shown by a brief resume of a rather complicated genealogy resulting from cooperation between the government and several industrial concerns. Whittle's first engine, the W.1, was designed by the company specially set up for that purpose — Power Jets, Ltd — with the unit being built by the British-Thomson Houston Co, which also built the successor, W.2. Then, in April, 1940, the Ministry of Aircraft Production decreed that Power Jets should concentrate on research and development, and it accordingly turned the W.2 drawings over to the Rover Co for production.

The British-Thomson Houston unit, modified by Power Jets, became the W.2 Mark IV; while modifications by the Rover Co resulted in the W2B, from which was derived the River class engines produced by Rolls-Royce and also the 1-A, first produced in this country by the General Electric Co as sire of its I-16 series.

Rolls-Royce interest in the turbojet field dates back to 1938, when Dr Griffiths joined its staff, and by 1941 test units had been put in action to study various components. Meanwhile the company's production facilities had been made available to both Power Jets and Rover, and parts of the W2B were made for the latter.

First of the Rolls-Royce River class engines, designed in collaboration with Power Jets and based on the Whittle conception, was the Welland, which reached the test stage in late 1942. A centrifugal-compressor unit with ten reverse-flow combustion chambers, it had static sea-level thrust of 1,700 lb and a weight of 850 lb. Passing its 100-hr test in April, 1943, the Welland went into production in May, 1944, as the power plant of both prototype and first production models of the twin-engine Gloster Meteor, which was used against Nazi V-1 buzz bombs.

Meanwhile, Rover engineers had done preliminary work on a through-flow design, but this type of unit did not find favor with Power Jets because it required a longer shaft between compressor and turbine, calling for installation of a third bearing. However, Rolls-Royce — which in April, 1943, took over Rover development work — went ahead with the prototype, known as the W2B.26 and it first ran in May, 1942.

This unit, first of the Derwent series, had a thrust of 2,000 lb at 16,500 rpm and a dry weight of 850 lb. Utilizing a two-sided impeller and ten combustion chambers, the engine had an overall diameter of 42½" and length 101½".

This unit was followed by Derwent Marks II, III, IV, and V — the latter being described in detail, together with the record-breaking Gloster Meteor, in Aviation for January (Page 69). The Derwent I and IV each gave a 10% increase in thrust; the III was an experimental installation set up for boundary layer control studies.

Though basically of the same design, the V has an increased-capacity centrifugal compressor and nine instead of ten combustion chambers. Fuel control, too, differs. On the I, pump capacity was fixed and flow to the combustion chambers controlled by bypassing excess fuel. On the Derwent V, pump capacity is varied by means of an aneroid to reduce fuel supply at altitude. Another change on the V is that the discharge nozzle box at the rear of the combustion chamber is a casting rather than a fabricated structure, a change made to eliminate splitting and cracking of welding due to the high temperatures. And securing of nozzle vanes inside the nozzle box has been improved by using a. buttress type connection, which reduces differential expansion of vanes and casings. Another important difference is in the bearings, the V using roller instead of plain bearings.

By May, 1944, Rolls-Royce had developed a reduction gear which was installed on a Derwent engine for driving a propeller, this unit finishing test stand runs in Mar 1945 and beginning flight tests in the Gloster Meteor in September of that year — the first British turbine-driven propeller engine to fly. This power plant, known as the Trent, is strictly an experimental unit built to study prop-jet problems. Its total weight is approximately 1,500 lb, 250 lb each for the propeller and reduction gears and 1,000 for the engine itself.

Latest of the Rolls-Royce River class engines is the W2B.41 (B/41), called Nene, on which details have not yet been released. Geometrically very similar to the Derwent V, it develops static sea-level thrust of 5,000 lb at 12,400 rpm with a dry weight of 1,550 lb, and it has specific fuel consumption of 1.06 lb/lb thrust/hr.

Growing out of the Whittle design, but developing with basic differences from it, have been the turbojets developed by Maj Frank Halford and the de Havilland Company. It was in January, 1941, that Maj Halford, who had developed the Gypsy and Napier reciprocating engines, was brought into the gas turbine power plant field. His organization worked closely with de Havilland interests and, in March, 1944, became, without personnel change, the de Havilland Engine Co, Ltd.

Main difference between the Halford and other British centrifugal engines lies in the fact that Halford engines — the Goblin I and II — use a single-sided impeller. Throughout the development program, the Halford interests have felt that the single-sided impeller had several distinct advantages. The direct ducting of intake air, for example, means that air flow can be increased with relation to the plenum chamber arrangement. It gives high velocity entry without flow breakaway, reduces pressure losses, takes quite full advantage of ram effect, and simplifies installation. While a two-sided impeller has a greater intake capacity for a given engine diameter, the need for getting air to the rear face means a larger overall installation diameter.

Design work was started in April, 1941, with the first H.1 reaching the test stage a year later, and flight test stage in March, 1943, when two were installed in an early Meteor type. First flights in the de Havilland Vampire were made in September, 1943, and in the prototype Lockheed XP-80 in January, 1944. Its thrust was brought up to 3,000 lb at 10,500 rpm, with a total weight of 1,550 lb and specific fuel consumption of 1.233 lb/lb thrust/hr.

In Goblin II, the 31" diameter impeller with 17 vanes delivers air to the combustion chambers at a 3.3:1 pressure ratio at maximum speed of 10,200 rpm, giving vane tip speed of 1.430 fps. One of the main differences between the Series I and II engines is in the design of the labyrinth seal on the aft face of the impeller, which matches with grooves on the sealing plate. In the latest series, a semi-buttress shape is used, compared with full buttress on the earlier model. Clearance of the lips on a cold engine is .015", with expansion under heat during operation bringing an overlap of .075"-.095".

Hollow rotor shaft, machined from a steel forging, is bolted to both impeller and turbine disk. The rotor assembly is supported on front and rear roller bearings, the former being mounted ahead of the impeller on a stub shaft and cooled by intake air, while the latter is mounted just ahead of the turbine disk and cooled by ducted air bled from the compressor between combustion chambers. No thrust bearing is required, since the single-sided impeller tends to pull forward, counteracting the pull aft of the turbine.

The single-stage turbine has 83 blades made of Nimonic 80, an alloy with a high nickel content. Blade roots are serrated "Christmas tree" type fitting into serrations on the turbine disk and held in place by peening the roots on both sides, the upstream peening being heavier to resist rearward thrust.

Goblin II has 16 flowerpot-type combustion chambers of three main elements: Outer casing, flame tube, and burner. The tapered cylindrical outer easing is deep-drawn mild steel, protected inside and out by nickel plating, fastened by a flanged joint to a die-cast aluminum expansion chamber (at upstream end) which is bolted to the diffuser casing. The downstream end is attached to the turbine nozzle junction assembly by a piston-ring type expansion joint.

The flame tube is concentric with the outer easing, held in place by three pins attached to the outer casing but permitted to slide inside sockets for radial expansion. The burner comprises outer and inner cup-shaped caps and flared cover plate.

Approximately 20% of the air entering from the compressor flows through a metering annulus into the inner cap for primary combustion, some going through a swirler around the fuel nozzle, the remainder around the swirler to give turbulence to mix fuel and air. The remaining 80% of the air, as it passes through the annular space between the outer casing and flame tube, is brought into the flame tube through ports.

Combustion takes place in the forward third of the chamber, with the air being added serving to reduce temperatures from 2,000°C near the burner to not more than 790°C at the turbine.

Another highly interesting British development is the Bristol Theseus, a project now in the experimental stage and on which little information is available. Designed for use on long-range medium-speed transport craft, the unit's propeller is driven by one of three turbines, and a heat exchanger, for heat regeneration from exhaust gases, is also used. As reported by Dr Harold Roxbee Cox, wartime chairman of the Jet Collaboration Committee, the Theseus comprises an axial-cum-circumferential compressor which delivers air through the heat exchanger to eight combustion chambers, then through three turbines, two of which drive the compressor and accessories, the third driving a conventional propeller through epicyclic reduction gear. In its present state of the unit's development, approximately 80% of the power is delivered to the propeller, the remaining 20% being used for jet.

All the units discussed, it will be noted, have been centrifugal-compressor types growing out of the original Whittle design. But work on axial-flow types has not been entirely neglected in England since Dr. Griffiths' pioneer work. For example, an REA 1938-designed 8-stage compressor known as Anne, was built in 1939 by Armstrong Siddeley and tested the following year. And in 1939 the Parsons organization built an 8-stage compressor called Alice.

As a result of these and other studies, Metropolitan-Vickers in 1940 began construction of an axial flow prop-jet, designated the D.11 and designed for 2,000 bhp. This project was halted, however, to concentrate on an axial-flow turbojet, and in July, 1940, the company started design work on the F.2, a 9-stage axial-flow compressor annular-combustion chambered 2-stage turbine unit. Two of these engines were installed in a specially modified Gloster Meteor, and the first successful flight of a British axial-flow turbojet was made on Nov 13, 1943.

Development on this project has continued to the point where the current unit, the F.2/4 with a 10-stage compressor and single stage turbine, delivers some 3,500 lb static sea-level thrust for a dry weight of 1,700 lb. This thrust is about twice that of the original unit, but the weight and overall dimensions — 42" diameter and 13' 3" length — are very little more than those of the prototype.

In 1942, Armstrong Siddeley was given its first gas turbine contract, the results being the ASX and ASP, the latter essentially the ASX driving a propeller.

The ASX was running within nine months, according to company, delivering its calculated thrust of 2,550 lb with specific fuel consumption of 1.0 lb/lb thrust/hr. The prop-driving ASP delivers some 3,600 shp and 1,100 lb thrust.

This basic plant is a reverse-flow axial unit, the air entering through ports about midway along the engine between the combustion chambers, turning 90° into the 14-stage axial compressor, then turning 180° into the eleven combustion chambers and on out through a two-stage turbine, as shown in the accompanying illustration.

This article was originally published in the March, 1946, issue of Aviation magazine, vol 45, no 1, pp 78-81.
The original article includes 4 photos, a drawing and a diagram of jet engines: Illustrations are not credited.